WhatsHap

WhatsHap phases genomic variants by assembling haplotypes from sequencing reads to assign heterozygous single nucleotide polymorphisms (SNPs) to the two copies of a diploid genome.

Key Features:

• Read-based phasing: Uses sequencing reads for phasing instead of relying solely on statistical population-based methods.
• Haplotype assembly: Assigns heterozygous SNPs to their respective haplotypes in diploid genomes.
• Weighted Minimum Error Correction (wMEC) solver: Implements an optimal solver for the weighted minimum error correction problem.
• Runtime scalability: Achieves runtime linear in the number of SNPs.
• Fixed-parameter tractability (FPT): Uses coverage as the FPT parameter to control computational complexity.
• Coverage handling: Efficiently handles datasets with coverages up to 20×.
• Practical coverage guideline: Demonstrates that ~15× coverage is generally sufficient to reliably phase long reads even with elevated sequencing error rates.
• Error-aware phasing: Accounts for sequencing error information and read length when constructing haplotypes.
• Accuracy metrics: Produces haplotypes with favorable switch and flip error rates compared to state-of-the-art statistical phasers.

Scientific Applications:

• Population genetics: Enables haplotype-resolved analyses required for population genetics studies.
• Downstream haplotype analyses: Provides phased variant data for downstream analyses that depend on accurate haplotypes.
• Long-read sequencing phasing: Applicable to phasing long-read sequencing data where longer reads and higher error rates are present.

Methodology:

Performs read-based phasing by optimally solving the weighted minimum error correction (wMEC) problem with an algorithm that is linear in the number of SNPs and fixed-parameter tractable using coverage as the parameter, while incorporating sequencing error information and read length.